Abstract: Alcohol-related motor-vehicle fatalities are no longer declining and are even increasing in the United States. It has been suggested that this stalling is partially due to inefficiencies in current policies, in particular the failure of current alcohol policies and prevention efforts to reach some vulnerable groups. Alcohol policies, such as per se laws, have been enacted based on decade-old estimates of the relative risks that drivers face. Some groups may be at a higher risk (e.g., females). Clear relative risk estimates for some relevant groups of drivers are unknown (e.g., repeat DWI offenders). Ongoing debates (such as those involving the reduction of the permissible BAC threshold from .08 to .05 or to efficiently allocate resources between specific deterrence and general deterrence programs) will benefit from this study. This proposed study will take advantage of a unique data set recently made available to provide more targeted information about the vulnerability of different groups of drivers to alcohol-related crashes, identify groups increasingly at risk, and estimate the maximum number of crashes that could be avoided if policies targeted to these groups were implemented. Among the groups of special interest are drivers aged 18 to 20, drivers with prior DWI convictions, drivers who are female, drivers at .05dBACd.08, drivers with a passenger, drivers with passengers aged 15 and younger, and drivers of racial/ethnic groups other than Whites. This information should be useful to policymakers in designing more efficient, better targeted countermeasures. PUBLIC HEALTH RELEVANCE: PUBLIC HEALTH RELEVANCE STATEMENT: This study will identify groups currently at risk and estimate the maximum number of crashes that could be avoided if policies targeted to these groups were implemented. This information will help to clarify ongoing policy debates and, therefore, will be useful to policymakers in designing more efficient, better targeted countermeasures.